When companies feel stressed
How a cost-cutting strategy led to the greatest collaboration effort of all time
In light of the economic turmoil following the pandemic and war, many companies are in the cost-saving mode. In the software sector, these are the headlines just in the last few days:
Oracle gears up to lay off thousands
Microsoft sells its lay off as reorg
Facebook (being nasty as usual) hides the lay offs behind performance reviews
Tesla being Elon’s show, lays off due to his “super bad feelings”
Ford lays off 8K
Google? keep reading…
Google has been a respected leader in the software industry for a long time. This internal memo from their CEO reveals some interesting bits:
Even a giant (both in terms of number of employees and revenue) is going into cost saving mode. The sh*t is serious! 😬
Although they're slowing down recruitment due to "uncertain global economic outlook", when they do hire, the focus is on "engineering, technical and other critical roles". Phew! 😅
Disclaimer: this post is by no means an endorsement of Google or Sundar Pichai. The focus here is on how cost reduction brings focus to a company.
"Scarcity breeds clarity"
Apart from slowing down hiring, Google asks employees to "consolidate where investments overlap and streamlining processes". This is a call to focus "pausing development and re-deploying resources to higher priority areas".
In the memo, Sundar Pichai says “scarcity breeds clarity” which really resonates with the story I am about to share.
Days of glory
Back in 2016 I joined a media company. Little did I know, I was part of a new product and tech organisation which was created solely to compete with Google and Facebook. The reason? Media traditionally relied on advertisement revenue but as Google and Facebook knew way more about people, they could deliver targeted advertisements. It was a more effective way to reach consumers and yielded a higher bang for buck. They ate the media's lunch!
But this media company wasn’t going down without a fight. Equipped with advice from McKinsey and enjoying deep pockets at the time, it created a new product and tech organisation and filled it with some of the smartest people I had the honour to work with. It was there where I first met an actual xoogler (ex-Googler) in the flesh! There were many, including the CTO. And a few from Facebook, Spotify and other media companies.
The opportunity of defeating the enemy is provided by the enemy himself.
Fast forward a few years and it became apparent that a media company focusing on the small Nordic countries (Norway and Sweden) is no match for American giants all while traditional media was losing popularity. The company broke in two halves and the CEO took the “better” half with him. You see, this wasn’t your average media company. It had diversified its investment by buying many startups, most of which were in the so-called “classified” market (basically like eBay or Craigslist). Most of those brands went to a new company led by the original CEO. The media part (where I was working) remained and got a new CEO.
Disclaimer: this is my recount of the story written in retrospect. Just to clarify, I wasn’t in “the room” where these decisions were made.
It took a few months for the new CEO to find the ropes. Then came a relatively benign announcement about some reorg in her management team. To us engineers, it wasn’t particularly big news. CEO could do whatever she pleases. I recall diversity and inclusion was a big theme at the time, more on that in another post. We didn’t care too much about those high level changes. If anything, we were eagerly waiting for more.
And more came: over the next few months the CEO reports started doing renovations of their own sub-orgs and the changes eventually trickled down to our level.
The unwritten strategy
Throughout my career at that company I remember “innovation” being a constant motto. In one form or another, it was everywhere, at every high level message, motivation and even the company slogan! I didn’t see much of it in the old days before the company split. We were just creating “yet another…”: yet another login screen, yet another CMS, yet another ad inventory, yet another news site.
I remember seeing a pair of Oculus Rift VR headsets at the office one day and tried them for the first time. What an amazing experience. When I asked “Why do we have these?” , I was told “to experiment with new ways of telling stories”. Yet by the time I left, I didn’t see any VR content on our news sites.
The failure of the new product and tech org and the ensuing collapse of the company was as much an external factor (traditional media loosing popularity, intense pressure from Google and Facebook on the ad revenue) as it was internal. This was a classic example of “culture eats strategy for breakfast”. So becoming a smaller company and focusing on media was a step in the direction that matched our culture.
In a truly Scandinavian spirit, each team enjoyed a high degree of autonomy. I guess you know where I’m going with this: We were heavily siloed. Each team had their own tooling, process, plans and (depending on who you asked) different definitions of the company mission and vision.
I wouldn’t go as far as calling it “total chaos” because somehow, through the good will and sweat of engineers implementing e2e features across teams, it worked. Those few engineers cross pollinated between the silos. I was one of them and eventually became a staff engineer. It wasn’t a smooth ride though. The end users would feel the consequences (like having a completely different theme between the product of the team which made the login screens, the teams rendering the news and the teams rendering the ads) but it wasn’t like the site was broken!
Like I said, it makes sense in retrospect: the new product and tech org which I joined a few years back was filled with smart people and those people had high salaries. When it was evident that we lost the ad war to Google and Facebook, there was an exodus of talent.
It’s a lie to say I wasn’t speculating to leave but the vacuum that was left by some of my smart ex-colleagues provided a great opportunity to learn and grow so I stayed a enough to witness the effect of the pandemic.
As many business sectors were going down one by one, there were whispers of a lay offs. One thing I really liked about the new CEO was her monthly live all-hands which ended with questions and answers. We used Sli.do and some used it to ask any questions anonymously, and the CEO answered brutally honest. Thanks to those monthly all-hands, we had a good alignment with the CEO. It helped the entire company to get the same memo. It greatly facilitated our collaboration across teams because we were sailing towards the same north star.
There was a certain uncertainty and sadness in the air when the CEO announced the loss of our first colleague to COVID-19. He was a photographer for one of the media brands owned by our company. She also confirmed the rumors of the lay off while promising that she’ll do everything within her power so it doesn’t come to that.
This was serious. The message was clear: now that we could not increase our ad revenue, we had to cut the operation costs.
I work with strategies these days. A proper strategy has a clear diagnosis paired with a guiding policy and coherent actions. At least that’s the format I operate upon based on the Good Strategy Bad Strategy. I didn’t see a single document with those elements, yet the entire company seemed to be aligned with the CEO. The unwritten cost-saving strategy turned into a great opportunity for people across the board to bring their best and collaborate.
Smash them silos
I worked at the Core Platform team. Not to bore you with the details, we made a white label news site platform which could be used by any of the 20+ brands the company owned. We were at the epicentre of the action (or so it felt) but every direction we looked, we saw silos.
We were sandwiched between the brands pursuing their own objectives and the upstream API teams which didn’t give what we needed in a timely manner due to being too distant from the end users. This was a symptom. The root cause? We were fragmented by tech, priorities, concerns and expertise. “You cannot collaborate when you don’t coordinate”, I often used to say in frustration.
Gradually, the cost-saving strategy smashed those walls. It started with an initiative to consolidate our observability tooling which gave us great visibility into what’s running across our stack. Not just our stack, but the entire stack that is involved from when a reader goes to the URL of a news article all the way down to the metal. This involved DNS, CDN, web, apps, BFF, tracking, ads, API and a bunch of microservices all the way to the database.
We crossed that data with our cloud bills to spot opportunities for cutting costs. Over the course of the next 12 months, we killed many servers, rewrote some to serverless and re-architectured part of the stack.
The company terminated the expensive contractors and we, the regular employees, had to fill in. They also deprecated the specialized QA or DevOps titles, which forced us, engineers, to put the operation hat and fully own the quality aspects of the code. There was an uproar internally, but as the time passed by, what didn’t kill us made us stronger. Personally, I learned a lot about operations and cloud computing during that period and took multiple courses on AWS and learned Terraform, Python and Golang although my main title was “frontend developer” and up until that job, I had always written the code and thrown it over the metaphorical wall for the “DevOps” (operations in actuality) to handle. The new “generalist profile” that they grew us into, was indeed very empowering. Not only I could touch anything from the CI/CD pipeline to the integration tests and performance monitoring, I was expected to. Hell I had admin access on the most critical accounts behind the face of the company. My entire team did!
I’m sure the end users saw the benefit too because the sites became snappier as we did less sh*t behind the scene. By the end of that year I and my colleagues Emil and Martin practically had the FinOps hat on.
I remember that I provisioned a lambda for a hack day project and 2 weeks later, Emil was tapping my shoulder to shut it down. This would save us a grand total of 2.1 cents per month! We were savages like that!
What’s more, we started collaborating with other teams to find what tech stack they are running and whether there are different solutions to the same problem.
Previously our collaboration stopped at the API level. Now, we were looking behind the scene.
Sure enough we found a bunch of opportunities, which were mostly remnants of the good old autonomous days of that McKinsey-inspired product and tech org. We killed them when we could and rewrote them when we couldn’t. It’s not like we had anything better to do.
Our primary objective was to keep the lights on and the heads above the water. This was existential. We were determined to cut the costs or a lay off would be inevitable.
Sidenote: a Swedish lay off is not as dramatic as it sounds. There are tons of safety nets in place to make sure you don’t end up sleeping in the streets in the Nordic winters. The strong presence of unions also makes sure that companies don’t fire people out of spite.
The lay offs didn’t happen, but that period of repeated investigative work eventually turned from a fun adventure to a routine chase-and-kill mechanical chore. There was no sign of light at the end of the tunnel either.
Eventually I decided that I’ve had enough of FinOps and got a full-time position as a SRE at a global media company. Emil and Martin quit within a few months after me.
The story didn’t end bad though: the sites are still up and despite us three removing our salary burden from the company, we all got higher salaries offsetting the impact of the inflation on our own pocket! win-win!
When the body is in stress, it goes into fight and flight mode. Adrenaline and cortisol increase the blood pressure to the muscles while limiting growth. It makes evolutionary sense. There’s no reason for the body to spend resources on growing your nails when you’re running from a tiger. Longer nails do not impact the tiger’s appetite! 🐅
When Sundar asks Googlers for "pausing development and re-deploying resources to higher priority areas", he’s sending clear stress signals. When he’s saying that new recruitment is limited to "engineering, technical and other critical roles", he’s talking about pumping blood to the muscles.
Alas, just as the body cannot sustain long periods of stress, a company cannot be in this mode forever either. People with a growth mindset would eventually pack their bags and try their chances somewhere with more interesting challenges.
There is one thing that hit too close to home in Sundar’s memo to Googlers: focusing on “streamlining processes” and reducing duplication of effort is the smart thing to do, fight and flight mode or not. That’s one of the areas good staff engineers can make an impact, but it's up to everyone to flag fragmentation and collaborate to increase efficiency.
Let’s end with the final paragraph from Sundar’s memo:
Scarcity breeds clarity. It’s what drives focus and creativity that ultimately leads to better products that help people all over the world. That’s the opportunity in front of us today, and I’m excited for us to rise to the moment again.
Update: Also read Layoffs Don't Tell the Whole Story
Thanks for reading Alex’s Newsletter! Subscribe for free to receive new posts.